Slashdot Mirror

Bit-rot is an issue inherent to any storage medium

Here's a quick article which explains how hard disks use error correcting codes so that the user-level experience is no bit rot but rather many many read failures before even a single block of undetectably corrupted data. Next time you can know what you're talking about.

http://www.pcguide.com/ref/hdd...

And no, HDDs do not have built-in error correction, they have checksums -- those things are not the same thing.

Sorry to inform you that your knowledge on this subject is not perfectly correct and inclusive. Hard drives use per-sector ECC. ECC stands for Error Correction Code. The very term tells you its function is to do precisely what you say is not done. Here is one tutorial. This stuff is pretty basic and widely known.

"When a sector is written to the hard disk, the appropriate ECC codes are generated and stored in the bits reserved for them. When the sector is read back, the user data read, combined with the ECC bits, can tell the controller if any errors occurred during the read. Errors that can be corrected using the redundant information are corrected before passing the data to the rest of the system. The system can also tell when there is too much damage to the data to correct, and will issue an error notification in that event. The sophisticated firmware present in all modern drives uses ECC as part of its overall error management protocols. This is all done "on the fly" with no intervention from the user required, and no slowdown in performance even when errors are encountered and must be corrected."

http://www.pcguide.com/ref/hdd/perf/raid/levels/multLevel15-c.html

Buy lots of hard drives and build a NAS using RAID 51. Pretty much never have to worry again. Use PAR2 if you want just in case files somehow get corrupted.

It's a way overboard solution, but it's pretty close to bulletproof. Using something like Carbonite or one of the other remote online storage methods mentioned would be much similar without the ridiculous costs a home RAID 51 would incur.

That's not radiation, that's because cheap CRTs tubes oscillate at 60 hertz and if you're not deaf in the upper frequencies you can hear them whine. Basically it's noise from the flyback transformer in the CRT. Many children can hear them but people often lose those frequencies as they get older.

I can still hear when a cheap CRT is on but I don't claim to be allergic to wifi.

See http://www.pcguide.com/ts/x/comp/crt/failWhine-c.html for some more info.

Hi MR AC! The problem is the MFT bites you in the ass every single time without fail. I have a friend that works in the state crime lab and it is ALWAYS the MFT that gets you. It is almost impossible to remove file traces from the MFT (I know of a few cleaners that will, but the vast majority won't touch it for fear of making the machine unbootable) and they can tell what you had on the drive simply by doing a search of the pointers in the MFT.

Now since I doubt he is using an OS where it is trivial to clear files from the file system like FAT (they also have tools to hunt for files in EXT 2/3 and ReiserFS) and in all likelihood he is running XP or later, well then the odds he could pull off erasing the MFT without making it obvious the MFT was tampered with (remember willful destruction is a felony IIRC) are pretty much zip. It is always the MFT that gets them, especially since so few understand how the MFT works.

http://www.virtualdub.org/blog/pivot/entry.php?id=71

http://www.pcguide.com/ref/hdd/file/ntfs/relFrag-c.html

http://www.virtualdub.org/blog/pivot/entry.php?id=71

http://www.pcguide.com/ref/hdd/file/ntfs/relFrag-c.html

that was already how they determined processor speed: http://www.pcguide.com/ref/cpu/char/mfg_Rating.htm

Those are "logical" sectors, which can be different from the physical sector size. [...] ...logical sector size is a drive interface level concept distinct from the filesystem cluster or block size. Filesystem block sizes have generally been larger than the logical or physical sector size for quite some time.

Thank you, butlerm for some much needed sanity in this thread.

The notion of cylinder/head/sector in the literal sense has been completely deprecated, but the scheme was preserved as an ad-hoc standard... which has since run into five other "hard limits" in BIOS and Int0x13 addressing... but that's another story.

Around the time of the first BIOS addressable-space limitation of 528MB, most hard disks were already being mapped with ZBR and translated through firmware. (see Zoned Bit Recording) The standard of 512b/sector has simply been a case of tradition and best practice. It was "just they way they made them."

A new logical int0x13 hook driver is all that's needed to interpret C/H/S coordinates with a 4KB base instead of a 512B base, and M$/Apple/Linus (et al) can likely cook that up in their sleep. No applications—short of low-level virus scanning, low-level disk utilities and software RAID, to name a few—would be affected by a different "physical" sector size. Most apps are in "virtual mode" and treat files as objects, which is handled in turn by the core OS.

Most filesystems use 4K granularity as it is, all that needs to happen is to equate cluster with sector. (as a most simplified "patch") I'm sure they'll come up with a new scheme to keep it scaled up... such as 4 sectors per cluster. (becoming 16K cluster/block size)

The greatest headache will be the game of catch-up by the OS-dev and utilities arena, who now have to completely review their preconceptions about HDD storage. Think regression-testing-cubed.

While we're at it, let's correct TFP by saying that filesystems will always align partitions to the beginning of the next logical cylinder. (whether it's a multiple of 4KB or not) The most likely problem to arise would be with the utilities mis-reading the "Advanced Format" disks, (showing them at only 1/8 of their actual size) calculating capacity based on total-sectors*512b, instead of the correct formula of total-sectors*4096b.

If you think about it, it's really a practical step. Think of where HDD sizes were ten years ago; 40GB was gi-normous and 100GB was just a pipe-dream. Ten years before that? A 100MB disk was "spacious" ...just a 1GB disk wouldn't be available for another five years. In all that time, the sector size hasn't changed at all... just half a kilobyte. It's a sign of the times, people.

Think of it this way... now instead of the nerdy term "sectors," you can introduce a newer term like "quads" or "kilo-quads" and see if it catches on.

Uh, no.
JBOD is when you link all disks together.
JBOD means "just a bunch of disks" strung together.

http://www.pcguide.com/ref/hdd/perf/raid/levels/jbod.htm

Could be more expensive than it's worth for the drive maker to build. They'd rather sell more drives for RAID10. Since the exact same drives could be sold to those who don't want the extra performance. Thus they get economies of scale.

Whereas it seems (even from the responses here) that most people are quite happy with the current drive sequential speeds, so they wouldn't pay extra for a larger multihead.

Similar reasons why multiple independent heads haven't taken off:

http://www.pcguide.com/ref/hdd/op/actMultiple-c.html

I wouldn't recommend using rotating-platter drives for "long term" archival storage, if by "long-term", you are talking more than 5 years. The lifespan of modern drives is such that 5 years is possible, but no guarantee.

The choice of format certainly won't be a factor in the ease of retrieval in the next 5 years, and probably won't in the next 10. ATA 1 was approved as standard back in 1994, and most motherboards today still have an onboard connector that supports it. That's what, 15 years of complete support? Any reason to believe that they will stop selling usb adapters for ide anytime soon? If you're concerned with sata's lifespan, just buy the $20 usb-to-sata adaptor now and you're covered. And realistically, what other drive interface would you suggest for home backup? SAS? Sata is the cheap, reliable and fast option these days.

The big issue with USB is that it is too slow for practical backups when the data involved approaches 1TB. With most USB external drives, you'd be lucky to break 20MB/s for sustained transfers. That's 50 seconds per GB, or almost 2 hours to backup the lot. The original poster was describing his effort to backup his media collection, so assuming that his backup needs involved hundreds of gigabytes of data is not too far fetched. Sata transfer rates on modern drives should be 50-60MB/s sustained, and quite possibly higher. This gives you a backup time of less than an hour, which is a lot more practical.

The external usb drive option also has a higher cost. The Freeagent unit adds about a $50 premium for each extra drive, not reasonable for a single unit, but quite expensive for a good swap set of 4 or 5 drives, and an unnecessary duplication of equipment. I'm not knocking the concept, but in practice, the external USB units are not as cheap or fast as a mobile tray and standard sata drives.

One could make a case for external drives with esata support, but you'd still be looking at an extra cost of $50 per drive, and a duplication of hardware. No thanks.

The great thing about this concept is that once solid state drives get to the point where they are cheap enough, they are a drop in replacement for the rotating platter drives you used to use. Instant upgrade in both reliability and speed. Assuming of course, that some manufacturer gets a clue and releases said drives in the 3.5" form factor. Don't get me started...

Conventionally, clock rates, bitrates, bandwidths, and other networking and data rates are understood to use decimal prefixes (i.e., 1 kb/s = 1000 b/s), RAM is understood to use binary prefixes (i.e., 1 MB/MiB = 1024 kB/KB/KiB), and storage devices use both in different places (the OS or the packaging).

Random sources that back this up:
http://www.speedguide.net/read_articles.php?id=115
http://www.pcguide.com/intro/fun/bindec-c.html
http://www.cknow.com/refs/BitsBytesandMultipleBytes.html
http://en.wikipedia.org/wiki/Bit_rate#Prefixes
http://en.wikipedia.org/wiki/Binary_prefix#Usage_notes

And do you know that current Intel and AMD processors emulate x86?

yeah right. What is this emulation you speak of?

The emulation that has formed the basis of all Intel's x86 processors from the Pentium Pro/Pentium II onwards, that's what!

The core of the processors are RISC-like (unlike the "genuine" x86 CISC cores of the original Pentium and older chips). x86 instructions are converted internally to RISC format before execution. (More).

To the best of my knowledge, this is transparent to the rest of the system and any code running on it. In fact, that's an interesting question- is there *any* way of accessing the CPU at the RISC level? My guess is that there must be, because IIRC Linux (and possibly Windows) can alter the behaviour of the CPU to compensate for bugs.

The instruction set has an awful lot of bearing performance. The RISC ARM instruction set has only instructions that take exactly 1 cycle (last time I looked). This makes both efficiency and optimisation such as pipelining very effective. The CISC x86 instruction set has instructions that can last varying amounts of time. This makes things such as branch prediction misses expensive. To compensate for this x86 chips use a translator which turns the x86 into VLIW pseudo-RISC internally. Unfortunately this translator takes up most of the power and silicon real estate on the chip.

You can compare the instruction sets of ARM and x86. Writing assembler for the former is a dream, unlike the latter. Mhz for Mhz, ARM blows x86 away.

Phillip.

http://www.pcguide.com/ref/cd/constSpindle-c.html

It's all from the audio CD days, being as they were forced to use CLV for those drives they didn't develop the technology to read CAV so it's easier to use what you already make and will have to continue to make rather than having two separate designs, as most computer CD players also play music CDs, one of which would have to understand both types of data transfer.

i believe the main problem associated with these ridiculously large storage capacities is that the exponential growth of storage capacity has been faster than the exponential growth of bandwidth.

1.5TB is the largest hard drive that isn't overly expensive, which (at the 1.9Mb/s bandwidth which is the median internet speed in America, is 76.6 days. Even in Japan (with a median of 61Mb/s) it takes 58 hours.

20 years ago with a 56k dialup and a 300MB hard drive it would only take 12 hours to copy it.

(hard drive capacity from http://www.pcguide.com/ref/hdd/hist-c.html , http://www.speedmatters.org/document-library/sourcematerials/sm_report.pdf , http://en.wikipedia.org/wiki/Modem and a few other places for an estimate of connection speed, i might be a little high for the historical estimate here)

I don't see this situation improving any time soon.

That's not quite accurate. The company that bought/took the formula was a company that manufactured electrolyte, not a capacitor company. That flawed electrolyte was sold to about a dozen capacitor manufacturers in Taiwan and Japan, who in turn sold the capacitors to probably thousands of companies. It affected products by almost every major computer vendor, including every manufacturer you mentioned in your post.

IBM
http://www.theinquirer.net/en/inquirer/news/2003/06/22/ibms-capacitors-are-dodgy-too-but-its-hushed-up

HP
http://news.cnet.com/PCs-plagued-by-bad-capacitors/2100-1041_3-5942647.html

Apple
http://news.cnet.com/Apple-offers-repairs-for-problem-iMacs/2100-1041_3-5841331.html
http://discussions.apple.com/message.jspa?messageID=2071244

While we're at it, Dell, Asus, MSI, Shuttle, ECS, Giga-Byte, Abit, and Compaq.

I doubt you can find any computer or motherboard vendor that didn't get bitten by those capacitors on at least one of their products.

That said, I do agree that buying from a major manufacturer is probably a good idea. The advantage of the more reputable, bigger name vendors is that when bad things happen that are outside their control (as this clearly was), they are more likely to stand behind their products even for people who didn't buy the extended warranty....

Intel already tried this on the Pentium processors, see here or here or here I would post more, but google is a great resource

Yes, I suspect that the author's idea of the "standard" keyboard layout everyone is expecting (*) is the 101 keys AT keyboard... that came later than all the keyboards of the list...

IBM PC keyboards
Comments on the layout

And I also find more and more difficult to find a keyboard with the standard 104-keys layout, or better. I have nothing against more key if they're not getting in the way (I used unix keyboards with more than 120 keys that were usable), but most keyboard models you can find now have annoying "features".

Examples :
- F keys default to multimedia keys, there's a fn key misplaced (on one model, right in the place of the left ctrl, which in turn is shifted to the right, and you get a smaller space bar
- delete key in place of ins + del
- power keys moving down the 6 keys editing block so "up" and "end" are next to each other
- editing block as 3 rows x 2, unusual and cramped layout, sometimes the left key does "cross" under the shift key, with a shorter right ctrl

Not about the layout, but there are also more and more "slim" keyboard using the laptop technology. Standards full stroke keyboard have a way better feedback.

(*) which isn't even an argument to say that a different layout is inherently bad

Actually the supreme keyswitch technology is electric capacitance such as that found on the Happy Hacking Professional keyboard from Fujitsu. The keyswitch passes between two metal plates to discharge an electic current that signals a key press. No physical contact takes place nor is necessary so you will have the longest, most consistent life and best responsiveness. This is usually backed up by a metal spring technology (such as the IBM Model M) to give tactile feedback.

If the Optimus doesn't use electic capatitance then it hardly deserves to be the "premium" keyboard. Then again I think the whole OLED thing is stupid and keycaps should be blank except for the bumps on F and J (on QWERTY). Print on keycaps is a crutch and the solution isn't to make better print! People just need to learn how to type. Computers should come with blank keycaps and an on-screen keyboard map so people don't get in the habit of looking at the keyboard and instead learn to look at the screen.

If your system still supports PS/2 and you want The Original then I advise looking into getting an older Model M from Clicky Keyboards. I got one of these for about $50. I would get a Customizer from Unicomp if my system only supported USB or I wanted to future proof myself against buying another keyboard in the future. Even then the Happy Hacking Professional is just too tempting.

Here's a summary of what changed between various versions of Windows, also taking into account that Windows 2000 is in a different product line than Windows 95/98/ME.

Home:
1995 - Windows 95 - New Windows GUI, new API, Plug and Play. Device Drivers are implemented directly in Windows instead of in DOS.
1997 - Windows 95b - Preliminary USB and FAT32 support, not available in stores
1998 - Windows 98 - USB and FAT32 support in mainstream. Unlike Windows 95, supports drives larger than 32GB. First version to include IE as a core component.
1999 - Windows 98SE - Windows 98 with a service pack. Seriously, there's no reason to buy it if you already have 98 as the updates are already on Windows Update.
2000 - Windows ME - Windows 98, but crashier thanks to the weirdness that was ME's System Restore. Tries to hide DOS and claim it isn't there, even though it is. Was quickly replaced by Windows XP.
2001 - Windows XP - Finally uses the Windows NT kernel. DOS dependence is gone. Much more stable than 98 or ME. First truly multi-user home version of Windows. System Restore actually kind of works, although I personally don't trust it.

Business:
1996 - Windows NT4 - First version of Windows NT to use the 95 GUI. Still no PnP support, though.
1996-1999 - Various NT4 Service Packs. No idea what they contained, I never used NT4.
1999 - Windows 2000 - Significant upgrades to the way NTFS and drivers work. PnP finally supported (and the crowds rejoiced).
2001 - Windows 2000 SP2 - Something major must have changed, because all modern software that runs on Windows 2000 requires SP2.
2001 - Windows XP - Windows 2000 with a new GUI and changes to the default IDE driver. ASPI is conspicuously absent, breaking CD Writing software that worked fine under 2000.

Other updates not mentioned above include Windows 2000 SP4 and XP SP2 (added support for larger ATA addressing to get around the ATA 137GB bug)

SO wrong on so many levels. Raid 1 does speed disc access: http://www.pcguide.com/ref/hdd/perf/raid/levels/si ngleLevel1-c.html
Slows write a little, speeds up reads.

This machine was out 2 months before Dell bought Alienware, I believe acer makes them.

Also it depends on what you consider your average dell lappie. The Inspirons are complete plastic crap you are correct, the XPS and Latitude line are quite nice, you get what you pay for...

Really, rather than an a whole new file system, I'd rather have native support for EXT3. Yeah, I know about ext2ifs. But I'd really rather install windows on an EXT3 partition, rather than being stuck with NTFS or a FAT32 partition arbitrarily limited to 32 GB. All this makes multibooting a PITA. (Or they can open up the NTFS spec so I can read/write in linux, but we all know it'll be a cold day in hell before that happens.)

Under NTFS, the maximum size of a partition (volume) is in fact 2 to the 64th power. This is equal to 16 binary exabytes, or 18,446,744,073,709,551,616 bytes. Does that seem large enough for your needs? :^)

http://www.pcguide.com/ref/hdd/op/actMultiple-c.ht ml

They want the simple cheap solution.

Sorry, but you seem to have raid's 5 and 10 reversed in your post.
Only raid 0 is faster than 10 (without proprietary equipment). Raid 5 is quite slow and 6 is even slower, especially for writing.

http://www.pcguide.com/ref/hdd/perf/raid/levels/mu ltLevel01-c.html

general RAID info (pictures)

http://www.acnc.com/04_00.html

http://www.pcguide.com/ref/hdd/perf/raid/levels/in dex.htm

Sorry, but you seem to have raid's 5 and 10 reversed in your post.
Only raid 0 is faster than 10 (without proprietary equipment). Raid 5 is quite slow and 6 is even slower, especially for writing.

http://www.pcguide.com/ref/hdd/perf/raid/levels/mu ltLevel01-c.html

general RAID info (pictures)

http://www.acnc.com/04_00.html

http://www.pcguide.com/ref/hdd/perf/raid/levels/in dex.htm

Via wanted to call a chipset the KZ, but KZ stands for "Koncentrationslager"

http://www.pcguide.com/art/rwtiu0006Chipset-c.html

So they renamed it to the KT. Luckily for them Tyrannosaur Americans haven't survived in sufficient numbers to form an effective lobby group, since they would no doubt find the acronym KT very insensitive.

Windows users get a key that does the same as right clicking (the Context key, seen here on the right. If you want to use a single button mouse you can, but quite why the hell you would want to is beyond me!

I agree completely with this. I had an end-user send a props email to my boss just because I explained the complexities of email systems to her using a cell phone analogy. She was so grateful to have someone explain a computer system to her in a way that really helped her to understand it.

PC Guide is an excellent example of this. It's somewhat out of date, but the author breaks down every component of a computer AND uses analogies and terms that Joe User could understand. It's pretty much how I taught myself about memory...among other things.

Another good example is How Stuff Works. They have an entire section on computers.

Hardware VS Software Raid

The $13 card you purchased is software Raid. Promise cards are mostly hardware RAID. I recently purchased a Promise FastTrack S150 SX4-M for less than $100 hardware RAID5 card compared to the $30-50 software RAID5 cards. I'm pretty satisified with the purchase but unfortunately there isn't room for much upgrade. I currently have 4x160GB in a RAID5 configuration giving me 480GB of space and 1 disc of redundnacy.

Some useful links to tell you the difference between software raid and hardware raid are:

http://www.pcguide.com/ref/hdd/perf/raid/conf/ctrl Hardware-c.html
http://www.redhat.com/docs/manuals/linux/RHL-9-Man ual/custom-guide/s1-raid-approaches.html
http://techrepublic.com.com/5100-10880_11-5715216. html
http://www.linuxplanet.com/linuxplanet/tutorials/4 349/2/

Psst... Floppys use FAT-12. http://www.pcguide.com/ref/hdd/file/partSizes-c.ht ml

Actually, Windows is a lot better than it used to be, in the old Win98/DOS days. You can do simple loops with commands like for (though they're ugly). To get help on any command, just do command /? (like -? in *NIX). There is help on some command-line commands in the Windows help (though it's hard to find).

As far as I have heard though, there have been also cases where SCSI have failed also.

Well, of course. There is no such thing as an infinite lifespan for hardware. Everything fails. It could be due to manufacturing defects, poor design, out-of-range operating conditions, or the device simply exceeds its designed lifespan and wears out. You cannot prevent all hard drive failures with better hardware. What you *can* do is eliminate the possibility of single-point failures. By using a RAID setup, or even simply a nightly off-site backup, you can ensure that more than one system would have to fail to cause data loss. Sure, it's possible, but statistically it's never gonna happen. It really would not be difficult for PC manufacturers to put two hard drives in a computer instead of one, and RAID0 them. Then the software would tell you when one of the drives failed. Thus the whole system is *about* to fail, and you haven't lost your data yet. But the customer base hasn't expressed a need for that, so computer manufacturers don't worry about it.

In any case though, we lack of a good way of predicting hard drives that are about to "die".

Yes, we do, because that adds unnecessary overhead to a system that is already a bottleneck in most systems (data i/o). How much notice would you like? 10 minutes? A week? A year? I can give you notice right now: your hard drive will fail sometime in the next 100,000 years. Sometimes it's just really hard to tell when a hardware failure is going to occur, and resource-intensive prediction software doesn't solve the problem.

In car services for example, nowadays, the plug the laptop on your car and they execute a full check on your car's condition. Picture this in hard drives. Then I would seriously consider buying a hard drive like that, even if the price was slightly above normal.

News flash: almost all hard drives manufactured since 2000 do this already. It's not a foolproof system, but neither is the computer system in your car, which failed to notify you that you are about to get rear-ended by that guy on the cell phone, thus causing a catastrophic car failure. Not all failures are predictable.

Because in doing so, I just found an explanation with a citation of someone actually trying this, and even a quasi-related previous Slashdot story.

Rule #1: Google it!

A filesystem format ("high-level" format) of a hard drive is much different than a low-level format.

Even a worthless A+ certification will teach you that much.

You do not even need to zero-fill the hard drive (which is NOT a low-level format) to wipe out a virus. If you reformat it at the OS level (again, a "high level" format) you're removing the FAT/MFT/inodes/btree pointers to the files on the disk, and without those pointers the data is pretty much inaccessible by the operating system, unless there is software in place to specifically go to each sector and read it - and because you've removed the pointers to the files that can kickstart that process when you do the reformat/reinstall, there is no way in hell that any such hidden virus code is going to be executed - you've removed the possibility.

Also: resorting to reformat/reinstall at the drop of a hat is a sure sign of incompetence. $.02 and then some. SOME spyware or viruses which include rootkits or other exploits are so intrusive that a reformat/reinstall may be recommended but it's not going to be every case. Hell, even Windows' "System Volume Information" (restore points) can be purged of infection by changing permissions on the directory and then scanning them- or you can even turn off System Restore to simply blow away the restore points and prevent automatic reinstallation of the scumware

Read. Learn. Stop scamming customers with technical terms you don't understand.

A filesystem format ("high-level" format) of a hard drive is much different than a low-level format.

Even a worthless A+ certification will teach you that much.

You do not even need to zero-fill the hard drive (which is NOT a low-level format) to wipe out a virus. If you reformat it at the OS level (again, a "high level" format) you're removing the FAT/MFT/inodes/btree pointers to the files on the disk, and without those pointers the data is pretty much inaccessible by the operating system, unless there is software in place to specifically go to each sector and read it - and because you've removed the pointers to the files that can kickstart that process when you do the reformat/reinstall, there is no way in hell that any such hidden virus code is going to be executed - you've removed the possibility.

Also: resorting to reformat/reinstall at the drop of a hat is a sure sign of incompetence. $.02 and then some. SOME spyware or viruses which include rootkits or other exploits are so intrusive that a reformat/reinstall may be recommended but it's not going to be every case. Hell, even Windows' "System Volume Information" (restore points) can be purged of infection by changing permissions on the directory and then scanning them- or you can even turn off System Restore to simply blow away the restore points and prevent automatic reinstallation of the scumware

Read. Learn. Stop scamming customers with technical terms you don't understand.

Far be it for me to diminish your story, but I did an F3 search for all my SID files (20327 of em, more files than any other music type) in the entire music directory, and it took about 15 seconds. Now that it's cached it takes about 3 seconds for successive searches on those directories. Most of the CPU time is spent in Explorer doing file name matching. dir \music\*.sid /s is similar. This is on an Athlon XP 2800, 1GB RAM, nVidia chipset, a 100GB NTFS volume and Windows Server 2003 sp1. Are you sure your filesystem isn't corrupted? What FS are you using, btw? You may need to defrag the MFT. When you say it hangs, do you mean the entire OS or just Explorer? Also, Explorer (shell) type searches will also search the index of ZIP files, so if you have a large or damaged ZIP file, Explorer might be choking on it. If you want the shell to quit doing that, unregister the zip folder support library:
regsvr32 /u %windir%\system32\zipfldr.dll

Basically the point of which is to remove a single point of failure (a single massive hard drive) from the server environment by allowing the data to be stored the same way in multiple places. There are 5 RAID levels, and the technology for RAID 5 (which is the highest level that I remember, anyway) has been out there and stable for a while.

It worked just fine, and would have continued to do so, but I decided to consolidate and move my services and firewall onto my main Linux workstation. It just didn't make enough sense to maintain and power a separate box for tasks that could easily be done with spare CPU cycles.

I still have the old chip and motherboard, though - it's special to me, as it was the first CPU I worked on out of school.

I've also heard that its more efficient to leave your computer on all the time because the amount it costs in wear and tear on your computer is more than the power saved by doing it. Anyone hear of this rumor or know anything about it?

Your history is a bit incomplete.* The EISA (1988) was the industries answer to the MCA (1987). Look at the card edge, and you'll notice it's basically the ISA (1985) bus with an additional row of pins. PCI came later (1994)

The VL-Bus was popular about '93-'94.

http://www.pcguide.com/index.htm

Raid 5 was specified.
That's what Raid 5 does.
Rebuilding a drive in raid 5 is a long intensive process, as it has to rebuild the parity.
Raid 5's often fail when a second near end of life drive dies while rebuilding a failed drive. The number one cause of this is heat buildup over time IMHO. Usually when a raid 5 looses 2 drives, it is actually the case that one drive failed, and there was no hot spare to take over, when the second drive fails (at some later date) the raid is broken. Anyone running a raid 5 with no hot spare is either an idiot or out of $$$.

http://www.pcguide.com/ref/hdd/perf/raid/levels/si ngle_Level5.htm

If you want more resiliance, go with multiple mirror raid1
Mirrors just duplicate the data exactly, which yields far quicker rebuilds.

Most people go raid 10 (Striping across mirror sets)

http://www.pcguide.com/ref/hdd/perf/raid/levels/mu ltLevel01-c.html

Raid 5 was specified.
That's what Raid 5 does.
Rebuilding a drive in raid 5 is a long intensive process, as it has to rebuild the parity.
Raid 5's often fail when a second near end of life drive dies while rebuilding a failed drive. The number one cause of this is heat buildup over time IMHO. Usually when a raid 5 looses 2 drives, it is actually the case that one drive failed, and there was no hot spare to take over, when the second drive fails (at some later date) the raid is broken. Anyone running a raid 5 with no hot spare is either an idiot or out of $$$.

http://www.pcguide.com/ref/hdd/perf/raid/levels/si ngle_Level5.htm

If you want more resiliance, go with multiple mirror raid1
Mirrors just duplicate the data exactly, which yields far quicker rebuilds.

Most people go raid 10 (Striping across mirror sets)

http://www.pcguide.com/ref/hdd/perf/raid/levels/mu ltLevel01-c.html

I wonder why you linked to Happy Hacking without mentioning their Blank Keytop model that more closely relates to the article. They even have it in Charcoal Black .

"Why does it cost $260?!" you might ask. Well the answer is that it uses capacitive keyswitches for both the greatest durability and tactile feel. They register key presses when a metal plate on the key moves between two parallel plates on the board that discharges a capacitive charge. Therefore there is no contact taking place to register the key press and therefore no wear on any contact points. To give tactile feedback each key has a standard mechanical spring mechanism (i.e. IBM Model M - note that Model M does NOT use capacitance, though).

There you have it. The ultimate luxury geek keyboard. It also has the CTRL key swapped with CAPS LOCK by default and the ESC key is where ~ would be on a US layout (to the left of '1'). A good bonus for UNIX nerds that have to work on a variety of boxen and don't want to setup the keyboard every time. Yes, hardware nerds can accomplish the same thing with some solder and wire. But for us software nerds that make a good salary it's a nice bonus.

2.5" are coming to the enterprise (Seagate Savio). Expect to see them in consumer gear too in a couple of years.

See pcguide for some reasons why the move to smaller platters is happening.